Final thesis for the obtainment of the M.Sc. degree (Master of Science)
Predicting the Consumption of Young Adults concerning Sugared and Sugar-free Beverages
Karen Dinkhoff University Twente
Niedenkampstr. 14 Communication Science
48565 Steinfurt Submitted to Dr. M.E. Pieterse
Tel.: 0172/2507861 and Drs. J.W.M. Verhoeven
Email: k.dinkhoff@student.utwente.nl
2 Acknowledgment
Firstly I would like to thank the agency fischerAppelt Kommunikation GmbH for the opportunity to develop this master thesis in the course of an internship. Furthermore, I want to thank them for their qualified support.
Special thanks to the Healthcare team for all information and to Mrs. Susanne Straetmans for her helpful supervision.
I would like to express my sincere appreciation to Dr. Marcel Pieterse and Drs. Joost Verhoeven who helped me with words and deeds during this master thesis.
3 Management Summary
In the course of an internship at the agency fischerAppelt Kommunikation GmbH a quantitative study was developed and accomplished with the aim to obtain information about the choice of non-alcoholic beverages made by young adults. On the basis of these findings, recommendations for an educational initiative were formulated, in order to reduce the uncertainty of the consumer and to provide a supportive help for making a responsible choice.
A written questionnaire, based on an adjusted form of the Theory of Planned
Behaviour from Ajzen (1991), was carried out by sixth formers and students. To measure the behaviour of young adults a Food Frequency Questionnaire concerning sugared and sugar-free beverages was developed.
342 sixth former and students, comprising both males and females, with the mean age of 20 years, completed the questionnaire during their lessons. On the basis of their results it was determined, which factors of the TBP have the most important impact on the choice of beverages and the intention to drink less sugared beverages.
The main conclusion of the study consists in the finding that the availability of sugared beverages at home proved to be the most important factor to predict the consumption of sugared beverages. The descriptive subjective norm of friends was the only variable out of the TBP which had some predicting power on the consumption of sugared beverages. The other constructs of the TBP turned out to be less influential. In particular, the correlation of the intention towards the consumption and the behaviour is unexpected low. Attitude, the injunctive subjective norm, the descriptive subjective norm of the parents, the self-identity concerning the health awareness and the perceived behavioural control were the components of the theory which account for the most variance of the intention. Moreover, female respondents and students showed a more positive attitude towards sugar-free beverages and consumed less sugared beverages than male respondents and sixth former.
These findings lead to some recommendations for the initiative in order to offer the consumer a supportive help to choose the right beverage. The initiative should focus primarily on the attitude. People’s attitudes can be changed due to comprehensible, individually relevant and credible messages, which include new arguments, but conform to the person’s existing perceptions and experiences. The initiative should include implementation intentions, which can be realized in everyday life. A successful approach to change old habits consists in presenting new habits which replace the old existing ones. In order to exert influence on the availability of beverages, it is necessary to present possibilities to the respondents how they are able to exercise control on the beverage purchase of the family.
4 Table of Contents
List of Abbreviations 6
Index of Tables and Figures 7
Foreword 8
1. Introduction 9
1.1 The Health and Eating Habits of Adolescents and Young Adults 10
2. Drinking und Eating Situation in Germany 11
2.1 Drinking Situation 11
2.2 Eating and Health Situation 12
3. Theoretical Framework 14
3.1 Behavioural Determinants 14
3.1.1 Biological Factors 14
3.1.2 Sociocultural Factors 15
3.1.3. Home Economical Factors 16
3.1.4 Psychological Factors 16
3.2 Social Cognition Theories 17
3.2.1 Previous Behaviour as an Independent Variable 20
5. Method 22
5.1 Development of the Questionnaire 22
5.2 Main Questionnaire 23
5.3 Procedure 26
5.4 Subjects 27
6. Results 28
6.1 Socio Demographical Factors 28
6.2 Consumption of Sugared and Sugar-free Beverages 28
6.3 Respondents’ Knowledge of Beverages’ Sugar Content and
the Daily Fluid Demand 29
6.4 Relations between the Variables of the Adjusted Theory of Planned Behaviour 30
6.5 Multiple Regression Analysis of the Beverage Consumption
and the Intention to Drink Less Sugared Beverages 33
6.6 Group Differences Concerning the Predicting Power of the Constructs 34
6.7 The Influence of Sociocultural Factors 44
6.8 The Influence of Home Economical Factors 45
6.9 The Influence of the Variables from the Theory of Planned Behaviour 46
5
7. Discussion 49
7.1 General discussion 49
7.2 Discussion of the used model 50
7.2.1 Influence of Knowledge 51
7.2.2 Influence of Attitude 52
7.2.3 Influence of Injunctive and Descriptive Subjective Norm 52
7.2.4 Influence of Perceived Behavioural Control 55
7.2.5 Influence of Self-identity 56
7.2.6 Influence of Intention 57
7.2.7 Influence of Previous Behaviour 58
7.3 Limitations of the Study 58
8. Recommendations 60
8.1 Methods to Increase Knowledge and Change Attitudes 60 8.2 Methods to Change Perceived Environmental Influence 61
8.3 Methods to Change Intentions and Habits 61
8.4 Differences between the Groups 62
8.5 Overview on the Most Important Advices 62
8.6 Scientific Recommendations 63
9. List of References 64
10. Attachment 71
6 List of Abbreviations
Abbreviation Meaning
TPB Theory of Planned Behaviour
KL Knowledge
A Attitude
ISN Injunctive Subjective Norm
DSN Descriptive Subjective Norm
SI Self-Identity
PBC Perceived Behavioural Control
I Intention
B Behaviour
7 Index of Tables and Figures
Figure 1: Adjusted Theory of Planned Behaviour 1
Figure 2: Stepwise Regression of Predicting the Intention to Drink Less Sugared
Beverages Regarding the Total Population (N=249) 35
Table 1: Frequency Table of the Socio-Demographic Facts 27 Table 2: Frequency Table of the Averaged Beverage Consumption per Month Ranked
by Mean Frequency 29
Table 3: Descriptive Table of Real and Estimated Sugar Content (100 ml/gram) 30
Table 4: Correlations of the Constructs 32
Table 5: Stepwise Regression Analysis of Predicting the Consumption of Sugared
Beverages Regarding the Total Population (N=249) 33
Table 6: Stepwise Regression Analysis of Predicting the Intention to Drink Less
Sugared Beverages Regarding the Total Population (N=249) 34 Table 7: Independent-Samples T Test of the Gender Differences 36 Table 8: Stepwise Regression Analysis for Predicting the Consumption of Sugared
Beverages Regarding Female Respondents (N=144) 37
Table 9: Stepwise Regression Analysis for Predicting the Intention to Drink Less
Sugared Beverages of Male Respondents (N=104) 38
Table 10: Stepwise Regression Analysis for Predicting the Intention to Drink Less
Sugared Beverages of Female Respondents (N=144) 39
Table 11: Independent-Samples T Test of Age Differences 40 Table 12: Stepwise Regression Analysis for Predicting the Consumption of Sugared
Beverages Regarding Sixth former (N=172) 41
Table 13: Stepwise Regression Analysis for Predicting the Consumption of Sugared
Beverages Regarding Students (N=76) 42
Table 14: Stepwise Regression Analysis for Predicting the Intention to Drink Less
Sugared Beverages of Sixth former (N=172) 43
Table 15: Stepwise Regression Analysis for Predicting the Intention to Drink Less
Sugared Beverages of Students (N=76) 44
Table 16: One-way ANOVA for the Educational Level of the Father 45 Table 17: Correlation Analysis between the Beverage Consumption and the Availability
at Home and at School / University 46
8 Foreword
Everyone has become accustomed to the great supply in every supermarket, has adjusted to it and cannot imagine abdicating it. A great supply of food and beverages comes along with being spoilt for choice, and therefore decision making is not always easy. In particular, the beverage market grows enormously. On the market the supply of beverages has become correspondingly diverse and therefore causes often confusion.
Besides this, the wide range of information concerning the fluid requirement complicates the customer’s choice of beverages. Some sources indicate that the recommendation of the daily intake of sugar-free beverages concerning adults should be at least 2.5 litres. In contrast, other sources present the advice of 1.5 litres fluid daily (Valtin, 2002; Negoianu
& Goldfarb, 2008). Many and partly contradictory information about juices, smoothies and soft drinks make the consumer insecure. Soft drinks and juices are often regarded unbalanced with respect to their ingredients, whereby the consumer has a guilty conscience due to the purchase and develops an “all-or-nothing” attitude. This leads to the questions: What kind of beverages and how much of these beverages should be consumed by the costumer?
9 1. Introduction
The diverse and complex information about beverages and their health benefits or threats confuse the consumer. A lot of consumers consider beverages as “allowed” and
“forbidden”, “good” and “bad” or “healthy” and “unhealthy”. This distinction of beverages cannot be applied so easily. Different studies show that inflexible diet-rules are not suitable for everyday life and that a violation of these strict rules leads to a loss of any control so that the forbidden food or beverage is consumed excessively (Ellrott, 2003).
To shed light on the uncertainty and confusion, which the consumer has to overcome almost every day, the agency of integrated communication, fischerAppelt Kommunikation GmbH, pursues the intention to instigate an educational drinking initiative for one of their customers. The planned campaign consists of three main goals: Firstly, it is aim to provide helpful recommendations concerning the drinking behaviour. Secondly, the campaign intends to offer a realistic impression of the current beverage market and to supply information about beverage’s ingredients. Thirdly, it is goal to raise the enjoyment in drinking beverages without overstraining the consumer and by presenting a great supply of non-alcoholic beverages to the customer.
To address the target group of the initiative properly more information is needed, for instance of their current drinking behaviour. The study compares the development of young adult’s preferences for sugared (e.g., juices, soft drinks) and sugar-free beverages (e.g., water, diet soft drinks). Sixth former (for a definition see § 5) and students allegorize the target group (young adults). It is characteristic for them to be highly educated and successful. Due to the fact, that they have to do intellectual work every day, they need a lot of energy to maintain their concentration. Therefore, the drinking and eating behaviour plays an important part because it provides them with the demanded energy. It is of high significance for the development of the campaign to investigate the group’s drinking behaviour in relation to their health awareness, their drinking motives and the needs of the educational advertising concerning beverages.
Only by surveying these kinds of information it is feasible to prepare the right advice and drinking information, which appeals to the consumers. The study tends to shed light on the factors, which determine the drinking behaviour. Why do some people consume more water and others more juices or soft drinks? Are there some factors which have an important explanatory power? On the basis of these findings recommendation for a campaign can be formulated, which help to reduce the uncertainty of the consumer and give supportive help for taking a responsible choice.
10 1.1 The Health and Eating Habits of Adolescents and Young Adults
The industrial countries observe a changing health situation of children and adolescents.
The German health care system has to deal with serious health problems, which affect a big part of the young population. For instance, more than 20% of German children and adolescents suffer from chronic diseases like asthma, diabetes or allergies (Warschburger, 2000). Furthermore, the life style of children and adolescents gets more complex due to growing leisure time facilities, an increasing food supply and due to changes in the family structures. Because of these changes and the missing comparable standards of older studies, the health behaviour was scrutinized by means of a new study, “the KIGGS” (Studie zur Gesundheit von Kindern und Jugendlichen in Deutschland) (Kurth, Bergmann, Dippelhofer, Hölling, Kamtsiuris & Thefeld, 2002).
KIGGS was set up by the Federal Ministry of Health and the Ministry of Education and Research and intends to collect a wide range of socio-demographic information and health related facts about the German youth. The goal of the study is to get a realistic picture of the health situation and in consequence of this being able to present advices for prevention programs.
An essential part of the health situation consists in the eating habits of adolescents.
In 2007 the results of the KIGGS-study concerning this area were published. The eating behaviour has, among other factors, a great influence on the development of diseases, like diabetes mellitus type 2, hypertension or cardiovascular diseases. Furthermore, the eating and drinking behaviour can predict our physical well-being und willingness to be active (Leitzmann, 2004). The body needs mineral nutrients, vitamins and fluid in order to function properly.
With reference to this awareness water and other non-alcoholic beverages play an important role in a balanced diet. The results of the KIGGS-study and other important studies concerning the health behaviour of young adults are discussed in the next chapter, following, an overview on the theoretical framework and the chosen theory (§3).
After this, the methods of the study and a description of the subjects (§4) are represented. On the basis of the theoretical background, the research questions are formulated (§5). In chapter six the results of the study are discussed, which provides a deeper interpretation of the discussion paragraph (§7). In addition to this, weak points of the study and possibilities of further research will be reviewed. Finally, recommendations for the current campaign are presented in the last chapter (§8).
11 2. Drinking und Eating Situation in Germany
2.1 Drinking Situation
The KIGGS-study indicates that 38.2% of the boys at the age of 14 to 17 years drink mineral water several times a day, whereas 46.7% of the girls drink this kind of beverage several times a day. 16.3% of the male probands drink tap water, respectively 21.2% of the female respondents. Likewise the drinking of mineral water, male adolescents consume soft drinks (36.0%). With the rising age the consumption of soft drinks increases significantly for both, boys and girls. However, male respondents are characterised by a considerable higher consumption of soft drinks in comparison to young females. The soft drink market has changed drastically in the last years and this has an effect on the fluid intake of water. The water intake of children and adolescents decreased between 1991 to 1998, whereas the soft drink intake increased (Sichert- Hellert, Kersting & Manz, 2001). The consumption of juices correlates negatively with the consumer’s age. The older the interviewees, the less juice they drink. Additionally, the study reveals that young males drink less tap water with increasing age. Another differentiation that can be made regarding the hydration is between migrants and non- migrants. Boys and girls with migrational background drink more often soft drinks. But this result is only significant for children and adolescents with a Turkish background and for migrants of German origin. Juices are more commonly consumed by non-migrants (Mesink, Kleiser, Richter, 2007).
To sum up, it can be observed that the consumption of soft drinks increases with age, the consumption of juices and milk beverages decreases and water consumption remains the same in the different stages of life.
The national consumption study of the Max Rubner-Institute (NVS), initiated in 2008, examines the eating behaviour and the health state of the German population at the age of 14 to 80 years. Women and men at the age of 19 to 34 years participate to a lesser extent in the survey. Thereby the percentage distribution of the study differs from the percentage distribution of the German population. Nevertheless, the study can be considered as an important foundation for the nutritional situation of young adults in Germany. According to the NVS, the recommended daily amount of 1.5 litres is accomplished by both genders so that hydration is satisfactory. The intake of water constitutes approximately half of the total hydration. Foremost, young men (19-24 years) consume a lot of water. Women at the age of 35 to 50 years consume the highest amount of water. A quarter of the total beverage consumption consists of coffee and green or black tea and 10% to 11% of the total composes of juices and nectars. 10% of the male respondents’ total fluid intake consists of soft drinks. This is twice as much as women’s amount of soft drinks (4%) referring to their total fluid intake. It turns out that
12 the socio-demographic status has an effect on the choice of beverages. It is scientifically proven that the lower social stratum of the population consumes sugared beverages, like soft drinks, three to four times more frequently than the upper social stratum (NVS I, 2008; NVS II, 2008). These facts are brought into focus in this master thesis because they are interesting analytical points for the curent study.
2.2 Eating and Health Situation
It is also crucial to scrutinize the eating behaviour of the German population and their assessments towards diets and the current health situation. 76% of all respondents opine that one can judge their health state as “very good” or “good”. One third (32.3%) of the consulted young men at the age of 14 to 24 years estimates their state even as “very good”. Women at this age share this view to a lesser extent (22.2%). The analysis of the food choice shows however that women eat more healthy products than men do (e.g., 270 gram fruit daily vs. 222 gram). It can be suggested that this may be similar for beverages. Women drink probably more beverages with less sugar and/or with additional health benefits.
More than half (52.6%) of the consulted respondents is not able to calculate their own energy demand. Foremost, women and men aged between 19 and 24 years state a negative calculation of their daily energy demand. Due to the fact, that people do not know how much they have to eat and drink every day, it can be expected that a lot of people eat and drink too much. Assigned to this study it can imply, that the respondents have a high sugar intake by consuming sugared beverages because they are unaware of their total energy demand.
The study shows that it is of high significance that women indicate a higher interest in nutrition than men do and that already young female adolescents aged 14 to 18 years take stock in this topic. Respondents with a higher graduation level (Abitur / Fachhochschule entry qualification) search for nutrition facts more often than people with a lower graduation (elementary school). To sum up, interest in nutrition correlates positively with education. And for this research one can act on the assumption that higher educated people have more interest and therefore more knowledge about healthy eating and drinking.
In spite of the interest for a healthy nutrition only 12.6% of the respondents accomplish the recommended daily vegetable amount of 400 grams. The German Nutrition Society (DGE) gives the advice to consume 250 grams of fruits every day.
However, only 59% of the population accomplishes this recommendation. Least fruits are consumed by men and women at the age of 19 to 24 years.
13 A study of the Federal Centre of Health Education (1992) examined the German population concerning their general food habits, their food consciousness and the appliance of different diets by observing their weight reduction. 52% of the young adults (20-29 years) declare that they pay attention deliberately to a healthy balanced diet. A study of Heseker, Adolf, Eberhardt, Hartmann, Herwig, Kübler, Matiaske, Moch, Nitsche, Schneider and Zipp (1994) reveals that young adults at the age of 18 to 24 years have an increased awareness of a healthy nutrition (cited in Allgöwer, 2000). According to a study, which is dealing with the eating behaviour of German students, 55.2% of the young consulted students attempt to eat healthy and balanced. Particularly, the female students pay a high attention to a balanced diet (Allgöwer, 2000). The health awareness concerning beverages is rarely studied. Because of this, the health awareness is part of the following analysis.
The outcomes show the Janus-faced side concerning the food situation of young adults. Oftentimes, there is a gap between the self-reported eating behaviour and the state of food intake. In the following passages, it will be clarified which determinants correlate with the drinking and eating behaviour and why many young female and male adolescents fail to apply a healthy nutrition in everyday life, although their strong will is existent.
14 3. Theoretical Framework
What predicts the drinking behaviour of young adults? Which influence has the sugar content on the choice decision of a beverage? To what extent can one relate the drinking behaviour to the general health awareness? The drinking and eating situation of young adults, in particular of students, is hardly studied in Germany (Allgöwer, 2000) and a complete overview cannot be given. Foremost, the drinking of non-alcoholic sugared and sugar-free beverages was not focused in any study. As a consequence, it is difficult to refer to studies concerning the drinking behaviour. Therefore studies examining the eating behaviour are consulted in the course of this work. The results of these studies will be presented as well as the relevance for the target group “young adults”. After the description of every determinant, a hypothesis is expressed with relevance to this study.
A distinction between sugared and sugar-free beverages is applied in every hypothesis.
3.1 Behavioural Determinants
The development of behavioural patterns is not that simple because a lot of influential factors interact with one another. One of these factors is for instance the social environment (Gedrich, 2003). In addition, behaviour can be developed on the basis of previous experiences, which demanded the same behaviour. One eats and drinks not just as a result of our biological requirements, but due to the influence of psychological, sociocultural and home economical determinants (Gedrich, 2003). In the next subparagraphs different influential factors are discussed, which might have impact on the decision making process.
3.1.1 Biological Factors
The genotype, in particular the gender, influences the eating behaviour. Although humans have an inherent preference for the taste “sweet” and an aversion towards
“bitter” (Ellrott, 2007), there are differences concerning the preference of taste between women and men. Empirical studies state that women eat more fruit, vegetables, milk products, whole meal-products and less sugared food than men do. These findings conclude that women eat healthier than men (Gerhards & Rössel, 2003). Other studies state that the age has an impact on the eating and drinking pattern, too. Older people show a healthier lifestyle pattern than young people (Rappoport, Peters, Downey, McCann & Huff-Corzine, 1993). Grimm, Harnack and Story (2004) affirm that the preference of taste has a big influence on what we drink and eat and is the most important predictor for the choice of a certain beverage.
15 Hypothesis 1: Female adults drink less sugared beverages than male adults.
Hypothesis 2: Older respondents (students) drink less sugared beverages than younger respondents (sixth former).
3.1.2 Sociocultural Factors
The predilection for a certain taste is not only based upon a genetic preference or a prenatal imprint but also on the determinant familiarity (Asp, 1999; Gniech, 2003). If something is frequently offered (mere exposure), the person becomes acquainted with this taste and is going to develop a preference for it (Contento, Williams, Michela &
Franklin, 2006; Gniech, 2003). This provides another socioccultural factor which can be defined as the influence of family and friends. Their drinking patterns affect the adolescent’s choice of beverages (Grimm, Harnack & Story, 2004). People identify themselves with certain groups and dissociate with others. If one decides to join a certain group, he or she is faced with different group norms, learns from other group members and adopts their behaviour patterns.
The sociocultural determinant education can be considered as a further aspect which leads to grouping of adolescents. People who have a higher education aim to be part of a special environment. Education is correlated with income and eating behaviour. People who belong to a better educational society often have more income and more money to spend for healthy food. Other explanations for this coherence can either be the knowledge of nutrition, which increases with the level of education (Kußmaul, Döring, Stender, Winkler & Keil, 1995) or the associated symbolic meaning of different foods. The more educated a social class, the more they associate fat and cheap products with a vulgar taste, which they dislike. Summing up, nutrition provides the possibility of choosing a certain group belonging (Barlösius, 1999).
The causal coherence between financial resources, education class and the choice of food cannot be indicated clearly. Whether a higher income leads to the purchase of more expensive, healthier and better food or a higher income results in a better education and therefore in more knowledge, is controversial up to now. But apparently one can only eat what one is able to buy. On the one hand, this is predicted by the income (Furst, Connors, Bisogni, Sobal & Winter Falk, 1996), and on the other hand by the availability of the product. Grimm, Harnack und Story (2004) and Gracey, Stanley, Burke, Corti und Beilin (1996) showed due to their research that accessibility, in particular the availability of soft drinks, water, juices and other food at their homes and at school, provides a significant influence on the beverage and food choice.
16 Hypothesis 3: Young adults have similar drinking patterns concerning sugar-free and sugared beverages as their parents and/or friends.
Hypothesis 4: Young adults from a higher educated family drink less sugared beverages than young adults from a lower educated family.
3.1.3. Home Economical Factors
Furthermore, the home economical situation determines the food offering at home. Who is responsible for the food? Which reasons specify the purchases? How are the food and beverages prepared and consumed? The values of eating and drinking are determined by the culture but can vary immensely in every family. People in a relationship eat healthier than single-living people (Roos, Lahelma, Virtanen, Prättälä, & Pietinen, 1998). This study also indicates that the presence of children has an impact on the purchase of food and beverages. Households with children younger than 14 years spend more money for healthy food than others households. Up to now, the mother is mostly in charge of purchasing and preparing the food, therefore she is the one who determinates which food is available (Gedrich, 2003). This leads to the hypothesis that children have similar eating patterns as their parents.
A study of Gracey et al. (1996) verifies that an unhealthy eating behaviour is caused by the inability of adolescents to exert an impact on the food choice at home. Apart from that, other studies have shown that healthy eating behaviour is associated with family structures and parents. Whereas the consumption of fast-food correlates with the relationship to friends, independence and fun (Backman, Haddad, Lee, Johnston &
Hodgkin, 2002).
Hypothesis 5: The availability of sugared beverages at home and at school has an influence on the consumption of sugared beverages among young adults.
3.1.4 Psychological Factors
Nutrition can be considered as a symbolic value. For instance, fast-food is associated with fun and in contrast, a familial setting leads to a consumption of healthy food among young adolescents. In this case, psychological processes play a significant role. Emotions, motives, attitudes and cognitive processes, like perception, thinking and learning, are parts of the psychological component.
Emotions excite a reaction in the individual, which can be either positive or negative.
Emotions – as a trigger for reactions – are not inevitably necessary for the food intake because food can cause a reaction without being triggered off by emotions namely due to the drive hunger. Nevertheless, some food products are able to evoke emotions.
Sometimes, one wants to eat or drink a special product just because it activates specific
17 emotions. Hence, it can be called a motif. By the choice of an aliment, people want to experience taste, put themselves in a certain emotional situation, express belonging or the social status.
Besides the factor emotion, the choice of food is determined by values, confidence, outcoming expectations, intentions and involvement (Gedrich, 2003).
The analysis of the influential factors shows that the determination of the eating and drinking behaviour is a complex process. It is impossible to analyse all determinants in one single study. Because of this, paragraph 3.2 gives an overview on two existing theories referring to the behavioural research, which can built a significant foundation for an empirical survey and the most important psychological factors (attitude, perceived subjective norm and perceived behavioural control).
3.2 Social Cognition Theories
Different behavioural models try to give theoretical answers. The theory of the “Health Belief Model” (Hochbaum, Rosenstock, Leventhal and Kegels, 1956) assumes that a behaviour depends on two variables: the value which implies how one credits the goal (e.g., to drink and eat healthy is important) and the appraisal which includes whether one can reach the goal by the current behaviour or not (e.g., eat fruit and vegetables).
The determinants which influence the behaviour are: the perceived susceptibility to develop a certain disease (diseases can be the result of an unhealthy diet), the perceived severity, perceived benefits or barriers and cues to action (Damoiseaux, van der Molen &
Kok, 1993).
Another explanatory model of behaviour is Ajzen’s “Theory of Planned Behaviour”
(1991). The “Theory of Planned Behaviour” supposes that behaviour is determined by the current behavioural intention. This intention is composed of three factors, namely the attitude of an individual towards an object, the subjective norm and the perceived behavioural control to accomplish a certain behaviour.
Attitudes result from certain motives and the positive or negative appraisal to what extent a certain object or behaviour is able to correspond with the individual motives (Backman et al., 2002). People, who have a positive attitude towards healthy food, are of the opinion that the consumption of healthy products leads to an improvement of their health, an increase of their energy and a good feeling about themselves. The chance that these people live a healthy lifestyle is considerably higher than for those people who have a negative attitude towards a healthy lifestyle (Gracey et al., 1996). Besides this, the study states that there is a relation between the positive attitude towards a healthy
18 diet and the desire to control one’s own weight, to lower cholesterol, to test their own willpower and to improve their appearance.
On the one hand, people know their own preferences and dislikes from personal experiences. On the other hand, people do not have to undergo these experiences but can basically develop own drinking and eating behaviour by observing others (imitation learning) (Bandura, 2001; Ellrott, 2007). The subjective norm refers to the individual perception, in particular, how important others value a certain object and how they behave.
The last factor, the perceived behavioural control, describes the own appraisal of the ability to accomplish a certain behaviour, although one has to overcome barriers and obstacles. The stronger the belief that one is able to eat healthy, the bigger is the chance to have a healthy eating behaviour.
In order to form an attitude, a subjective norm or a perceived behavioural control, it is necessary to receive knowledge of the topic. The numerous information people have to face every day, have to be selected, organized and interpreted. In this way a subjective perception concerning a healthy eating behaviour, which co-determines people’s (purchase) decisions, is developed. Studies show that the knowledge of health and a balanced diet has an influence on our eating behaviour. In view of these results, it can be assumed that people who know a lot about food, consequently eat and drink healthier (Backman et al., 2002). The ignorance of food ingredients can lead to a barrier, which hinders a healthy eating behaviour (Gracey et al., 1996).
The main aim of the study is not to built up a preventive intervention trying to persuade the target group to drink less sugared beverages. The study rather tries to get an overview of the factors, which influence the choice of beverages. Consequently, the TPB fulfils the requirements of the study to a greater extent than the other theory and is therefore applied.
Nevertheless, additions have to be made in the “Theory of Planned Behaviour” to improve its validity. The perceived behavioural control can be both, the control over external properties (e.g., availability) and the internal control over abilities and skills.
There are some studies which report that there has to be made a distinction between internal and external perceived behavioural control (IPBC, respectively EPBC), (Armitage
& Conner, 1999a; Conner & Armitage, 1998; Terry & O’Leary, 1995; White, Terry &
Hogg, 1994).
The second change, which has to be applied, concerns the subjective norm. On the one hand, a subjective norm can be developed due to the perceived values of other people and on the other hand, it can resolve from the perception of other people’s performances. This distinction can be subdivided into two different categories which are
19 named injunctive (“ought meaning of social norm”) and descriptive (“is meaning of social norm”) (Cialdini, Reno & Kallgren, 1990, p.1015).
The third and last addition to the already existent study is the construct self-identity.
Self-identity may be defined as “the salient part of an actor’s self which relates to a particular behavior” (Conner & Armitage, 1998, p.1444). Several studies from different health related areas have stated, that the performance of a certain behaviour is consistent with one’s own self-identity and thereby can be seen as an enrichment of the Theory of Planned Behaviour (Pierro, Mannetti & Livi, 2003; Sparks & Guthrie, 1998;
Armitage & Conner, 1999b; Sparks & Shepherd, 1992).
By relating the adjusted TPB with the complex of problems named above, one obtains the following structure:
The young adult’s knowledge of sugared and sugar-free beverages and their ingredients, as well as the arisen awareness of these beverages exerts an influence on the attitude. It can be assumed that someone, who deals with a healthy diet and consequently knows a lot about ingredients and beverages, has a more positive attitude concerning healthy beverages and therefore consumes less sugared drinks (Duffey &
Popkin, 2006).
The attitude, which is based upon knowledge, norms and values, predicts how a beverage is valued. If one emphasises strong flavour, there will be a bigger addiction to consume beverages which are more flavourful than water. If a greater importance is placed on drinking healthy beverages, the tendency will show a higher consumption of water.
The subjective norm is based upon the individual perception of the social environment and on the observation of other people. What does the environment consume? And how do they talk about beverages? If all friends drink soft drinks, the chance will increase that a young adult consumes the same beverage in order to belong to the certain group (Contento, Williams, Michela & Franklin, 2006). As named above, a clear distinction between injunctive and descriptive subjective norms (ISN and DSN) is applied in the study.
To what extent do young adults rate themselves as being able to drink healthy beverages at any time? Which obstacles are hit? And is it possible to control the obstacles? Theses answers can be measured due to the internal and external perceived behavioural control (IPBC and EPBC). The first question refers to one’s own abilities and skills and the other remaining questions advert to the external possibilities.
People create self-perceptions which are characteristics they ascribe to themselves.
In this study the self-perception is called self-identity. Young adults, who consider their lifestyle as healthy, show a healthier drinking behaviour. They basically drink water or tea in order to comply with their ideals (Sirgy, 1986).
20 The last construct of the theory is based upon the intention to perform a certain behaviour. If someone makes the plan to achieve something, the chance will be significantly higher, that he will actually fulfil the intention, than the chance for someone who does not aim to achieve the self-made plan. In the “Theory of Planned Behaviour”
the intention is the only direct link to behaviour and is determined by all the factors named above.
3.2.1 Previous Behaviour as an Independent Variable
Some studies discovered that previous behaviour belongs to the independent variables and has influence on the intention, which can be seen as the dependent variable (Towler
& Shepherd, 1991; Godin & Gionet, 1991). If someone always behaves in a certain way, the intention will comply with his or her behaviour, so that the person does not perceive dissonance between act and intention (Festinger, 1957).
In figure 1 the arrow between the variable behaviour and the variable intention reveals this coherence. The arrow has a broken line to clarify that it is not safe to say yet which connection exists. With the help of a regression analysis the relationship between intention and behaviour will be examined.
It is important to make the annotation that this study is a cross-sectional one. This indicates that the behaviour is measured only one time and therefore its constancy cannot be reviewed.
Hypothesis 6: The more adults know about the beverages’ ingredients the less sugared beverages they drink.
Hypothesis 7: The more positive the attitude towards sugared beverages the higher the consumption of sugared beverages (the worse the intention to drink less sugar-free beverages).
Hypothesis 8: The stronger the perception of the injunctive subjective norm concerning sugared beverages the higher the consumption of sugared beverages (the worse the intention to drink less sugar-free beverages).
Hypothesis 9: The stronger the perception of the descriptive subjective norm concerning sugared beverages the higher the consumption of sugared beverages (the worse the intention to drink less sugar-free beverages).
21
Behaviour
Intention
Attitude Perceived
Behavioural Control Subjective norm
Injunctive norm
Descriptive norm
Self-Identity
Internal PBC
External PBC Knowledge
Hypothesis 10: The worse the perception concerning the internal behavioural control of drinking sugared beverages the higher the consumption of sugared beverages (the worse the intention to drink less sugar-free beverages).
Hypothesis 11: The worse the perception concerning the external behavioural control of drinking sugared beverages the higher the consumption of sugared beverages (the worse the intention to drink less sugar-free beverages).
Hypothesis 12: The less the self-identity considering oneself as healthy, the higher the consumption of sugared beverages (the worse the intention to drink less sugar-free beverages).
Hypothesis 13: The lower the intention to drink sugar-free beverages the higher the consumption of sugared beverages.
Hypothesis 14: The higher the consumption of sugared beverages the lower the intention to drink sugar-free beverages.
With the aid of the following questionnaire, it will be examined to what extent the factors named above exert influence on the drinking behaviour of young adults.
Figure 1: Adjusted Theory of Planned Behaviour
22 5. Method
5.1 Development of the Questionnaire
The pre-test pursues the aim to test the comprehensibility of the questionnaire. This offers the possibility to exclude or change questions, which are difficult to understand and therefore mistakable questions and statements are avoided in the final survey.
To identify the factors, which affect the choice of beverages, an open questionnaire was sent to 30 German students by email (see attachment 1). Five students returned a completed questionnaire and on the foundation of these findings some additional answer possibilities were applied. Taste, healthfulness, a good value for money and the effect of the beverage (e.g., adipsous and refreshing) were the most important determinants named to choose a special beverage. The other questions are based upon studies from Armitage and Conner (1999a), Povey, Conner, Sparks, James and Shepher (1998) and Sparks and Shepherd (1992). The English questions were translated into German in order to avoid difficulties with the English language.
In addition, two different experts of the topic and people from the target group were asked to comment on the questionnaire with the aim to increase the internal and substantive validity of the constructs. In an open face-to-face questionnaire three staff- members of fischerAppelt were requested to check the questionnaire regarding formulations, content and structure. All of the respondents are very familiar with the topic of sugared and sugar-free beverages. With the help of their advices the questionnaire became more explicit and every important topic was considered. The respondents usually commented on the questions concerning the perceived behavioural control. Although the translation into German has the aim to avoid understanding problems, comprehension difficulties occurred because some items cannot be translated one-to-one.
Besides that, three people of the target group (two students and one pupil) were asked to comment on the comprehensibility and the content of the questionnaire. Their remarks were integrated in the ongoing questionnaire. Due to this applied pre-tests one can be sure that the questionnaire is coherent for the target group. Furthermore, the three consulted young adults expressed problems with the questions concerning the perceived behavioural control in almost the same manner. The given remarks led to following changes: The question “Ich habe viel eigene Kontrolle über das Trinken von zuckerfreien Getränken” was changed into „Es liegt in meinen eigenen Händen, wie viele zuckerfreie Getränke ich trinke“, the statement “Ich finde es schwierig, immer zuckerfreie Getränke zu trinken, da ich oft nicht weiß, wie viel Zucker in den Getränken steckt” was modified into „Ich finde es schwierig, immer zuckerfreie Getränke zu trinken, da ich oft nicht weiß, in welchen Getränken Zucker ist“, „Die Situation kontrolliert oft das
23 Trinken von zuckerfreien Getränken“ was changed into „Es ist oft von der Situation abhängig, ob ich zuckerfreie Getränke trinke“, „Das Trinken von zuckerfreien Getränken ist teuer“ was altered into „In Anbetracht des Geldes, das mir zur Verfügung steht, kann ich mir zuckerfreie Getränke nur selten leisten“ and „Es gibt Faktoren, die außerhalb meiner Kontrolle liegen, die mich am Trinken von zuckerfreien Getränken hindern“ was changed into „Es gibt Faktoren in meinem Umfeld, die ich nicht beeinflussen kann und mich deshalb am Trinken von zuckerfreien Getränken hindern“.
To test the substantive validity, two students absorbed in the study communication sciences were asked to read all questions without knowing, which question belongs to which construct. They were asked to assign the questions to the constructs, which fitted the different questions most suitable. Thereby it can be detected to what extent the
“measure is judged to be reflective of, or theoretically linked to, a construct under study”
(Farrell, Souchon & Durden, n.d., p.4). During this pre-test some questions were not assigned rightly but they were not changed because they were tested in the same way in other studies and there they turned out be questioned properly. This occurred especially within the construct of self-identity. The question “Es liegt in meinen eigenen Händen, wie viele zuckerfreie Getränke ich trinke” or „Ich fühle mich wohl, wenn ich zuckerfreie Getränke trinke“ were assigned to the construct self-identity. But these questions do not reflect an individual characteristic. It is rather an estimation of the person’s ability and feelings. Because of that, no changes took place.
5.2 Main Questionnaire
A questionnaire (see attachment 10.4), asking respondents about drinking sugar-free and sugared beverages, was constructed. At the beginning of the questionnaire, the topic of the study was explained and it was affirmed that the answers of the respondents are treated confidentially. The respondents obtained a detailed definition of sugared and sugar-free beverages and which beverages are not included in the survey. The questionnaire measured additionally the self-identity, internal and external perceived behavioural control, injunctive and descriptive subjective norms, as well as the central components of the “Theory of Planned Behaviour” (knowledge, attitude and intention) and moreover requested demographic information. The questions were presented in a fixed order and were mostly measured on a five-point Likert scale, which stretches from disagree completely to agree completely. The variables were reverse-coded, where it seemed appropriate.
Sixth former were asked to answer nine demographical questions and students ten questions concerning their social and demographical background (age, gender, size, weight, class, field of study, family background and educational achievement of the
24 parents). Two questions were applied to detect which beverages are available at their parent’s home and at school.
Knowledge (KL). The questionnaire included four questions about the knowledge concerning beverages and their contents (e.g. “Wie viel Gramm Zucker befinden sich in 100 ml (Angaben beziehen sich auf die gezuckerte Version des Getränks?”). Different possible answers were offered, including the right answer. To incorporate the difficulty, concerning the estimation of the beverages’ sugar amount every answer which deviated –3 points and +3 points from the officially right answer, was considered right. The three point variance was chosen because 3 grams represent one sugar cube and this is the established benchmark to rate the sugar amount of beverages. Finally, the right answers were added up, so that everyone was able to achieve a knowledge index between 0 and 15 points.
Attitude (A). 19 questions concerning the attitude (e.g., “Zuckerfreie Getränke sind langweilige Getränke”) were surveyed in the questionnaire. Most questions are based on studies of healthy eating behaviour, which used the “Theory of Planned Behaviour” (see § 5.1). Besides, different questions about beverages and their characteristics, which were developed from the pre-test, were applied. Both, direct and indirect measures of attitude, were used in the study. A semantic differential scale was used to assess the direct attitude of the consulted young adults. The participants were requested to respond to the sentence “Im Allgemeinen finde ich das Trinken von zuckerhaltigen Getränken…” on a five-point bipolar scale (e.g., “lecker” compared to “widerlich”). The indirect measure was composed of different statements, which the respondents had to rate on a five-point Likert scale from “Stimme überhaupt nicht zu” up to “Stimme voll und ganz zu”. The construct attitude was coded positively and reliable due to usage of the Cronbach’s Alpha coefficient .84.
Injunctive subjective norm (ISN). The injunctive subjective norm was tested due to four statements (e.g., “Meine Eltern finden es gut, wenn ich zuckerfreie Getränke trinke”).
Important people, who may influence the participants, are parents, friends and girl- or boyfriends. In this case an indirect measure with the five-point Likert scale was also used. The Cronbach’s Alpha coefficient of the summed scale was .75. This construct had a positive coding.
Descriptive subjective norm (DSN). Three statements were used in order to test the influence of the descriptive subjective norm on the choice of beverages (e.g., “Meine Eltern trinken mehr zuckerhaltige Getränke als zuckerfreie”). A five-point Likert scale was
25 applied to measure the answers of two questions. One question asked for the number of friends, who drink sugared beverages every day. The responded questions offered three types of social influencers: the parents, the partner and the friends. The construct had a negative coding concerning sugar-free beverages. The Cronbach’s Alpha coefficient was .52 which represented a low internal consistency. One can speak of an internal consistency if the alpha value lies above .70 (Spector, 1992). The influencers can have a conflicting drinking behaviour, therefore every person or group was respectively taken apart.
Perceived behavioural control (PBC). The analysis was carried out for the overall construct perceived behavioural control because the Cronbach’s Alpha coefficients for internal and external perceived behavioural control were rather low. The eleven items belonging to the two constructs resulted in a Cronbach’s Alpha coefficient of .60. After deleting one item („Zuckerfreie Getränke sind Zuhause häufiger verfügbar als Zuckerhaltige“) the coefficient was 0.61. The analysis of the factors released no explanation for the low consistency. Since other studies used the same items and proved the reliability of the construct perceived behavioural control with the existing ten items, it is used in the further calculations. The construct had a positive coding and the items were measured on a five-point Likert scale.
Self-identiy (SI). The respondents were asked to answer four questions concerning their self-identity based on a study of Sparks and Shepherd (1992) (e.g., “Ich sehe mich selbst als jemanden, der sich mit gesunder Ernährung beschäftigt”). All items regarding a healthy eating behaviour were measured on a five-point Likert scale ranging from
“Stimme überhaupt nicht zu” to “Stimme voll und ganz zu”. The answers referring to the self-identity resulted in a Cronbach’s Alpha coefficient of .42. The Cronbach’s Alpha coefficient was .62, after deleting the last item. This construct was used in several studies and therefore it is utilized further on. To examine this low reliability, an analysis of the different factors was arranged (see attachment 10.5). The first three items concerning the health awareness composed one factor. The other factor consists of the last item “Ich sehe mich selbst als jemanden, der Essen und Trinken genießt”. It was clarified that a distinction between the three items, which describe the health awareness, and the one item concerning the hedonism, is necessary. The construct describing the health awareness was coded positively and the single item concerning the hedonism was coded negatively.
Intention (I). Three items relating to the intention were presented on a five-point Likert scale by assessing the participants’ intention to drink more sugar-free beverages (e.g.,
